Three experiments were conducted to gather evidence for Wolford's (1975) feature perturbation model and claims derived from it. The first experiment found that perturbation or mislocalization of features toward the foveal center predominated when a distinctive standard for localization was presented and short-term memory factors were minimized. The second experiment found that foveal mislocalizations do seem to cause two figures in the visual field to be reported as closer to one another than they actually are presented. The final experiment provided no support for the notion that interference between figures in the visual field is caused by foveal mislocalizations making features from the figures appear to be at the same location. A distinction between mislocalizations of figures due to feature perturbations and errors due to figures which were correctly localized but not reported in the instructed order was possible in the last two experiments.